Varaksin A., Shalaumova J., Panov V. Control principles of confounders in ecological comparative studies: standardization and regressive modelss // Principy èkologii. 2014. Vol. 3. № 1. P. 4‒14. DOI: 10.15393/j1.art.2014.3341


Issue № 1

Analytical review

pdf-version

Control principles of confounders in ecological comparative studies: standardization and regressive modelss

Varaksin
   Anatoly
Institute of Industrial Ecology, varaksin@ecko.uran.ru
Shalaumova
   Julia
Institute of Industrial Ecology, yulyash@gmail.com
Panov
   Vladimir
Institute of Industrial Ecology, vpanov@ecko.uran.ru
Keywords:
confounding variables
accounting confounders
standardization
risk factor
analysis of observational data
regression models
Summary: The methods of the analysis of research data including the concomitant variables (confounders) associated with both the response and the current factor are considered. There are two usual ways to take into account such variables: the first, at the stage of planning the experiment and the second, in analyzing the received data. Despite the equal effectiveness of these approaches, there exists strong reason to restrict the usage of regression method to accounting for confounders by ANCOVA. Authors consider the standardization by stratification as a reliable method to account for the effect of confounding factors as opposed to the widely-implemented application of logistic regression and the covariance analysis. The program for the automation of standardization procedure is proposed, it is available at the site of the Institute of Industrial Ecology.

© Petrozavodsk State University

Reviewer: V. A. Ilyukha
Received on: 21 April 2014
Published on: 17 December 2014

Displays: 3336; Downloads: 579;